Guided by our ability to monitor in vivo growth, regression and relapse of prostate adenocarcinoma in mouse models, we are now well positioned to isolate tumors at specific stages of the disease progression. Considering the well recognized importance of stromal-epithelial interactions in prostate gland development and in prostate cancer, the central theme of this renewal application is formed on our desire to use this advantage of the model to begin to define important mediators of heterotypic cell interactions that may be critical in progression and recurrence of prostate cancer. All results will then be followed in human prostate cancer samples. Our hypothesis is based on the contention that the multiple secretory and inductive factors including bone morphogenetic proteins (BMPs) and BMP-induced activation of stromal cell-derived factor 1 (SDF-1), a novel interplay between prostate cancer cells and prostate cancer-associated fibroblasts (CAFs) that we have identified, serve critical autocrine, paracrine and endocrine functions in the regulation of growth and survival of prostate cancer cells and associated CAFs, in the intravasation and metastasis of the cancer cells, and in the angiogenic phenotype in the growth and progression of prostate cancer. For the recurrent or androgen depletion-independent (ADI) cancer that emerges following regression from androgen deprivation therapy, we hypothesize that additional changes in inductive factors and in the cellular compartments including epigenetically and/or genetically evolved CAFs may be important contributors. Our first aim is to elucidate the mechanisms by which BMP induces CAF cells to produce and secrete SDF-1, using cell systems from both mouse and human prostate cancer, and to define the biological effects of BMP-SDF-1 axis in prostate tumor progression. Our goal is to generate a mechanistic understanding by which BMP and SDF-1 mediate the reciprocal interactions between prostate tumor cells and non-tumor cells within the tumor. In the second aim we will characterize the progression of the ADI prostate cancer with respect to changes in the phenotypic distribution in the epithelial compartment, and with respect to potential molecular and functional changes in the mesenchyme, particularly CAFs that may influence the proliferation and progression of ADI cancer cells. Finally, the third aim concerns an exploitation of the study system to determine the origins of the ADI cancer cells. We will define the prostate stem/ progenitor cell subpopulations during prostate cancer progression, and characterize the functions of these subsets in the varied context of the coevolution of CAFs with the progressive changes in the neoplastic epithelial cells. A series of interconnected and parallel studies, using the mouse models, mouse prostate cells, human prostate cells, and clinical specimens is described to obtain mechanistic insight into specific heterotypic cell interactions that contribute to overall prostate cancer progression, metastasis and recurrence. PUBLIC HEALTH RELEVANCE: We have developed mouse models that mimic the progression of human prostate cancer. We now propose to examine the models to understand the mechanisms of cancer growth, metastasis, regression under anti-androgen therapy, and the recurrence of androgen depletion-independent cancer with the goal to identify new targets for prostate cancer therapy. .